Matches in SemOpenAlex for { <https://semopenalex.org/work/W4284975785> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W4284975785 abstract "Self-supervised skeleton-based action recognition with contrastive learning has attracted much attention. Recent literature shows that data augmentation and large sets of contrastive pairs are crucial in learning such representations. In this paper, we found that directly extending contrastive pairs based on normal augmentations brings limited returns in terms of performance, because the contribution of contrastive pairs from the normal data augmentation to the loss get smaller as training progresses. Therefore, we delve into hard contrastive pairs for contrastive learning. Motivated by the success of mixing augmentation strategy which improves the performance of many tasks by synthesizing novel samples, we propose SkeleMixCLR: a contrastive learning framework with a spatio-temporal skeleton mixing augmentation (SkeleMix) to complement current contrastive learning approaches by providing hard contrastive samples. First, SkeleMix utilizes the topological information of skeleton data to mix two skeleton sequences by randomly combing the cropped skeleton fragments (the trimmed view) with the remaining skeleton sequences (the truncated view). Second, a spatio-temporal mask pooling is applied to separate these two views at the feature level. Third, we extend contrastive pairs with these two views. SkeleMixCLR leverages the trimmed and truncated views to provide abundant hard contrastive pairs since they involve some context information from each other due to the graph convolution operations, which allows the model to learn better motion representations for action recognition. Extensive experiments on NTU-RGB+D, NTU120-RGB+D, and PKU-MMD datasets show that SkeleMixCLR achieves state-of-the-art performance. Codes are available at https://github.com/czhaneva/SkeleMixCLR." @default.
- W4284975785 created "2022-07-10" @default.
- W4284975785 creator A5030898980 @default.
- W4284975785 creator A5037378495 @default.
- W4284975785 creator A5046716542 @default.
- W4284975785 creator A5048813179 @default.
- W4284975785 creator A5057287134 @default.
- W4284975785 creator A5058973152 @default.
- W4284975785 date "2022-07-06" @default.
- W4284975785 modified "2023-09-27" @default.
- W4284975785 title "Contrastive Learning from Spatio-Temporal Mixed Skeleton Sequences for Self-Supervised Skeleton-Based Action Recognition" @default.
- W4284975785 doi "https://doi.org/10.48550/arxiv.2207.03065" @default.
- W4284975785 hasPublicationYear "2022" @default.
- W4284975785 type Work @default.
- W4284975785 citedByCount "0" @default.
- W4284975785 crossrefType "posted-content" @default.
- W4284975785 hasAuthorship W4284975785A5030898980 @default.
- W4284975785 hasAuthorship W4284975785A5037378495 @default.
- W4284975785 hasAuthorship W4284975785A5046716542 @default.
- W4284975785 hasAuthorship W4284975785A5048813179 @default.
- W4284975785 hasAuthorship W4284975785A5057287134 @default.
- W4284975785 hasAuthorship W4284975785A5058973152 @default.
- W4284975785 hasBestOaLocation W42849757851 @default.
- W4284975785 hasConcept C138885662 @default.
- W4284975785 hasConcept C151730666 @default.
- W4284975785 hasConcept C153180895 @default.
- W4284975785 hasConcept C154945302 @default.
- W4284975785 hasConcept C18969341 @default.
- W4284975785 hasConcept C199360897 @default.
- W4284975785 hasConcept C204321447 @default.
- W4284975785 hasConcept C2776401178 @default.
- W4284975785 hasConcept C2779343474 @default.
- W4284975785 hasConcept C41008148 @default.
- W4284975785 hasConcept C41895202 @default.
- W4284975785 hasConcept C70437156 @default.
- W4284975785 hasConcept C86803240 @default.
- W4284975785 hasConceptScore W4284975785C138885662 @default.
- W4284975785 hasConceptScore W4284975785C151730666 @default.
- W4284975785 hasConceptScore W4284975785C153180895 @default.
- W4284975785 hasConceptScore W4284975785C154945302 @default.
- W4284975785 hasConceptScore W4284975785C18969341 @default.
- W4284975785 hasConceptScore W4284975785C199360897 @default.
- W4284975785 hasConceptScore W4284975785C204321447 @default.
- W4284975785 hasConceptScore W4284975785C2776401178 @default.
- W4284975785 hasConceptScore W4284975785C2779343474 @default.
- W4284975785 hasConceptScore W4284975785C41008148 @default.
- W4284975785 hasConceptScore W4284975785C41895202 @default.
- W4284975785 hasConceptScore W4284975785C70437156 @default.
- W4284975785 hasConceptScore W4284975785C86803240 @default.
- W4284975785 hasLocation W42849757851 @default.
- W4284975785 hasOpenAccess W4284975785 @default.
- W4284975785 hasPrimaryLocation W42849757851 @default.
- W4284975785 hasRelatedWork W2043075591 @default.
- W4284975785 hasRelatedWork W2052253960 @default.
- W4284975785 hasRelatedWork W2375904069 @default.
- W4284975785 hasRelatedWork W2901298109 @default.
- W4284975785 hasRelatedWork W2944724518 @default.
- W4284975785 hasRelatedWork W2951391129 @default.
- W4284975785 hasRelatedWork W2970216048 @default.
- W4284975785 hasRelatedWork W3202244193 @default.
- W4284975785 hasRelatedWork W4225258897 @default.
- W4284975785 hasRelatedWork W4281640556 @default.
- W4284975785 isParatext "false" @default.
- W4284975785 isRetracted "false" @default.
- W4284975785 workType "article" @default.